Leak-Tightness in LPG Fittings: What Buyers Should Specify and Test

Premi Brasscom International has machined brass LPG fittings for over four decades, and in that time the single most common reason a fitting fails in the field is not the alloy or the thread form but a leak path that was never caught at the bench. This post is written for the buyer or sourcing engineer who needs to know what to specify, how to test, and which standards to cite when ordering LPG fittings in brass.

Quick answer: Leak-tightness in an LPG brass fitting depends on three things: a correctly machined thread with the right seal geometry, a torque value that compresses the sealing element without deforming the body, and a documented test method (helium or air-pressure decay, or a bubble test) performed on every part or on a statistically valid sample. Specify the standard, the test pressure, the allowable leak rate and the sampling plan in writing on the drawing or purchase order.

What causes leaks in brass LPG fittings?

Most leaks trace back to one of four causes: a thread that is out of tolerance (too loose to seal, or too tight to assemble), a sealing surface that was damaged during machining or handling, an O-ring or copper washer that was the wrong hardness or thickness, or an assembly torque that was too low to compress the seal or too high and cracked the body. Each of these is preventable at the machining and inspection stage, which is why a good supplier will quote a leak-test step as part of the process, not as an afterthought.

Thread form matters. Tapered threads (NPT, BSPT) seal by interference along the thread flanks and need a sealant or PTFE tape. Parallel threads (BSPP, metric) seal on a face or shoulder and need a copper washer, O-ring or bonded seal. Mixing the two without changing the sealing method is the most common specification error we see in RFQs.

Which standards govern leak-tightness in LPG fittings?

For LPG pressure vessel valves and fittings, EN 13175 defines design, material and testing requirements including external and internal leak-tightness tests. For vehicle LPG installations, BS EN 1949 requires each system to show no more than 5 mbar pressure loss when tested at 150 mbar air. For domestic appliance connections, BS EN 15069 covers safety gas connection valves for metal hose assemblies. If your market is India, IS 15100 and IS 14898 apply for LPG equipment.

Ask the supplier to name the standard on the quote and on the certificate of conformance. A certificate that says “tested per internal procedure” is not the same as a certificate that says “tested per EN 13175 clause 8.4”.

What test methods should a buyer specify?

Three methods are common, and each suits a different production volume and risk level.

Method How it works Best for Limitation
Bubble / leak-detection foam Pressurise the part, apply foam or soap solution, watch for bubbles Low volume, 100% inspection Operator-dependent, slow
Air-pressure decay Pressurise, isolate, measure pressure drop over a fixed time Medium volume, batch testing Sensitive to temperature changes
Helium mass spectrometry Pressurise with helium, detect helium with a mass spectrometer High volume, tight leak rates Equipment cost, needs trained operator

For LPG fittings, specify the test pressure (commonly 1.5x maximum working pressure, or a fixed value such as 1.5 MPa), the hold time (typically 10 to 60 seconds depending on volume), and the allowable leak rate (e.g. no detectable loss). For critical applications such as LPG cylinder valves, specify 100% inspection, not sampling.

What should a buyer put on the drawing or purchase order?

Write these five items into the specification or RFQ so there is no ambiguity at inspection time.

  1. Standard and clause: e.g. “Leak-tightness per EN 13175:2019+A1:2020, clause 8.4 (external) and 8.5 (internal).”
  2. Test pressure and medium: e.g. “Air at 1.5 MPa, held for 30 seconds.”
  3. Allowable leak rate: e.g. “No detectable pressure drop; no bubbles visible within 30 seconds.”
  4. Sampling plan: e.g. “100% inspection” or “AQL 1.0 per ISO 2859-1.”
  5. Documentation: “Certificate of conformance with test results attached to each batch.”

If you also need dimensional and material certificates, our quality page lists the documents we provide with each shipment, and our enquiry form is the fastest way to send a drawing for quotation.

Key takeaways

  • Specify the standard, test pressure, leak rate and sampling plan in writing on the drawing or PO.
  • Match the thread form (tapered vs parallel) to the sealing method; do not mix them.
  • Choose the test method to match your volume and risk: bubble test for low volume, helium for tight rates.
  • Ask for a certificate of conformance with test results, not a generic “tested” statement.

FAQ

What test pressure should I specify for LPG fittings?

For LPG fittings, a common specification is 1.5 times the maximum working pressure, or a fixed value such as 1.5 MPa (approximately 15 bar) for standard LPG service. The pressure should be high enough to reveal a leak path but not so high that it deforms the seal. Confirm the value with the supplier against the applicable standard.

Is a bubble test enough for LPG fittings?

A bubble test is sufficient for low-volume production and for fittings where a visible leak is the failure mode. For high-volume production or fittings with tight leak-rate requirements, an air-pressure decay test or helium mass spectrometry gives more consistent, operator-independent results and a documented leak rate.

What is the difference between external and internal leak-tightness?

External leak-tightness is the resistance to leakage from the fitting to the surrounding atmosphere. Internal leak-tightness is the resistance to leakage across the valve seal when the valve is in the closed position. Both are tested separately under EN 13175 clause 8.4 and 8.5 respectively.

Do I need 100% inspection or sampling?

For safety-critical LPG components such as cylinder valves and regulator bodies, specify 100% leak testing. For non-critical fittings in high-volume production, an AQL sampling plan (such as AQL 1.0 per ISO 2859-1) is acceptable, provided the supplier documents the plan and the results. When in doubt, specify 100%.

Scroll to Top